Development of New Pentaplex PCR Assay for Differentiating Staphylococci from Other Bacteria with Simultaneous Detection of Staphylococcus aureus Genes Encoding Panton-Valentine Leukocidin and Methicillin Resistance

In some human infections including those of blood, skin and respiratory tract, the causative bacterial agents tend to overlap, especially Staphylococcus spp. and Streptococcus spp. Such overlaps constitute difficulties in the choice of diagnostic tools, antibacterial chemotherapy and infection control strategies. To resolve this challenge, we developed a pentaplex PCR assay which simultaneously detects sequences for the recognition of bacteria (bacterial 16S rRNA), the genus staphylococcus translation elongation factor Tu (tuf), Staphylococcus aureus (spa), mecA-encoded Original Research Article Okolie and James; JABB, 2(4): 250-259, 2015; Article no.JABB.2015.025 251 staphylococcal methicillin resistance (mecA) and the S. aureus Panton-Valentine leukocidin (PVL) virulence factor (pvl). The new pentaplex PCR assay was validated using standard bacterial strains (N=377) including strains from the Network on Antimicrobial Resistance in Staphylococcus aureus (NARSA), the National Collection of Type Cultures (NCTC), and the National Collection of Industrial and Marine Bacteria (NCIMB). The new pentaplex PCR assay enables inference of bacterial presence/absence, differentiates between the genus Staphylococcus spp. and other bacteria, separates S. aureus from other Staphylococcus spp., differentiates between methicillin-susceptible and methicillin-resistant staphylococci, and detects the S. aureus PVL gene locus. The negative predictive value was 100% while the positive predictive value was 100%. Using a 96-well plate, the time to result was 2.5 hours against ≥24 hours by bacteriological culture. The new pentaplex PCR assay can easily be integrated into routine diagnostic microbiology workflow especially for laboratories with slim budgets which are unable to incorporate next generation sequencing at the moment.


INTRODUCTION
Recent studies have illuminated the abundance and diversity of bacteria in/on different parts of the human body [1,2].As more microbiomic data appear, differential diagnostic tools will be necessary to identify particular genetic factors associated with diseases and their outcomes [3].The first strains of methicillin resistant Staphylococcus aureus (MRSA) were reported in the United Kingdom hospitals about the same time as the introduction of the antibiotic methicillin [4,5].Following the UK reports, MRSA clones with diverse genetic backgrounds and virulence factors have been reported worldwide [6].Following the establishment of the centrality of mecA gene in the mechanism of staphylococcal methicillin resistance [7], polymerase chain reaction (PCR) detection of mecA DNA gained wide acceptance [8].
In 2004, it was observed that some MRSA strains were not detected by mecA PCR, rather, five primers specific to different staphylococcal cassette chromosome mec (SCCmec) right extremity sequences, including three new sequences and specific to the S. aureus chromosomal orfX, a gene of unknown function, located to the right of the SCCmec integration site, were the most reliable for the identification of some MRSA strains [9].The orfX information informed the development of several commercial MRSA assays including Xpert MRSA assay (Cepheid, California, USA) and BD-GO (Becton-Dickinson, USA).However, neither orfX nor the recently reported hybrids and analogues [10,11] reduced the centrality of mecA in staphylococcal methicillin resistance.
Whereas mecA gene driving methicillin resistance is shared among S. aureus and other staphylococci [6,7], Panton-Valentine leukocidin (PVL) encoded by lukSF-PV operon and staphylococcal protein A (Spa) encoded by the spa gene, both of which are associated with skin and airways diseases, are unique to S. aureus [12,13].The staphylococci and other bacterial species associated with human infections all share the tuf gene encoding translation elongation factor Tu as well as the 16S rRNA gene [14].Independently, S. aureus strains expressing Spa, PVL, or PBP2a are sources of concern to infection control, laboratory diagnosis and antistaphylococcal therapy [15]; their convergence highlights a higher threat to public health microbiology [16].It has been suggested that the application of multiplex PCR for the detection of multiple genetic markers within the same sample will provide a major contribution to the efficiency, logistics and cost-effectiveness of molecular diagnostics [17].In this regard, some previously developed MRSA and PVL assays have been revisited recently.As Pichon and colleagues envisaged that S. aureus harbouring mecA (LGA251) , a recently identified livestock variant of mecA, could be wrongly identified as methicillin-susceptible S. aureus (MSSA) in the absence of antimicrobial susceptibility testing (AST), they modified a triplex real-time PCR assay published in 2005 by McDonald and colleagues [18] into a quadruplex assay [19].
There are several reports of simultaneous sequencing and clustering of large numbers of 16S rRNA genes enabled by next generation sequencing (NGS) [20].However, multiplex detection of virulence and antibiotic resistance genes is not common at the moment.NGS machines and analysis of the associated metagenomic sequence data are currently possible only by high-budget laboratories [21].PCR-based MRSA and PVL assays do not distinguish between staphylococci and other bacterial agents of skin and airways disease.To provide this necessary advancement, we studied a triplex assay for simultaneous detection of S. aureus spa, mecA and lukSF-PV [22] and added two more DNA targets, namely: Bacterial 16S rRNA and staphylococcal tuf, thus arriving at a new pentaplex PCR.Sequencing and phenotypic tests were used to validate the new assay according to current practices.

Bacterial Strains and Microbiological Media Used for This Study
All microbiological media and consumables, except otherwise mentioned, were obtained from Oxoid (Oxoid, Basingstoke, UK).Bacterial strains were obtained from ultra-cold (-80°C) storage, sub-cultured in brain heart infusion (BHI) broth and plated out on BHI plates and sheep blood agar (Oxoid, UK) as reported previously [23].A total of 377 bacterial strains were used for evaluation through four panels (2.

Bio Safety
Safety is emphasised in our laboratory.Given that most NARSA isolates are potentially hazardous, work on them were carried out in enhanced level 2 (BSL2+) suite designated for such pathogens.Local bacterial strains were handled completely in the general (BSL 2) laboratory space where they have been circulating prior to this study.

Preparation of Bacterial Inoculum for PCR and Phenotypic Tests
For each bacterial strain used in this study, a 0.5 McFarland standardised inoculum was prepared and used for PCR and for phenotypic tests.

Extraction of Bacterial DNA for PCR
Extraction of DNA from NARSA isolates grown on BHI plates was performed by heating and centrifugation as reported previously [24].Briefly, 0.5 mL live bacterial suspension was killed (95°C, 10 minutes).Following a quick centrifugation (13000 RPM, 20 seconds), the DNA-rich supernatant was transferred into a fresh 0.5 mL Eppendorf tube for PCR in the general (BSL 2) laboratory.
DNA extraction was waived for local strains.Isolates were applied directly into the PCR from BHI broths, blood agar plates, BHI plates and from 0.5 McFarland broths.

Design of Oligonucleotide Primers
The sequences of the oligonucleotide primers used in the new pentaplex PCR assay are presented in Table 2.The primers targeting spa, pvl and mecA were published originally for a triplex real-time PCR assay [22].The primers targeting 16S rRNA and tuf were designed for this study.Numerous nucleotide sequences for each marker were sourced from the National Centre for Biotechnology Information (NCBI) and entered into clusta lW suite at the European Bioinformatics Institute (www.ebi.ac.uk/tools/clustalW) for multiple alignments which allowed identification of highly conserved motifs (HCMs) for hybridization of primers targeting bacterial 16S rRNA and staphylococcal tuf genes.Primers were located manually according to Mount [25].All primers were purchased from Sigma Genosys (Sigma, UK).

Performance and Optimization of New Pentaplex PCR
PCR amplification was performed incorporating 5.0 µL of the template in a 40 µL PCR using Eppendorf mastercycler (Eppendorf, Hamburg, Germany) following the cycling conditions described by Nakagawa et al. [22] with modifications.Briefly, an initial single cycle for 5 minutes at 94°C was followed by 40 cycles consisting of 15 seconds at 94°C (denaturation) and 5 seconds at 60°C (amplification) with one final run of 30 seconds at 72°C (final polymerase extension).The reaction was then cooled to 4ºC for post-amplification processes.PCR optimization followed previously described methods [26,27].The optimized factors driving the new pentaplex PCR assay are listed in Table 3.

Gel Resolution and Gel Purification of PCR Products
PCR products were resolved using conventional Tris-EDTA buffer submarine electrophoresis (200 V, 1 h) in 2% agarose gel containing Ethidium bromide (0.5 g/L) and visualised in UV transilluminator (UVP, UK).PCR products were purified from agarose gels using GenElute™ (Sigma, UK) according to manufacturer's instructions.

Validation of the New Pentaplex PCR Assay Using Conventional Phenotypic Tests
Without disclosing their genotypic and phenotypic properties, bacterial strains were taken from the freezer and tested on the new pentaplex PCR assay as RSS001, RSS002, RSS007, etc,.The finding of Gram positive cocci in clusters was used to characterise Staphylococcus spp.Tube coagulase test was used to differentiate S. aureus from CoNS.Staphylococcus spp.were examined for staphylococcal methicillin resistance by oxacillin salt agar screen (OSAS) according to Clinical and Laboratory Standards Institute (CLSI) recommendation [28] with modification of the oxacillin content based on a more recent work used in the UK to support the real-time PCR detection of mecA and mecA (LGA251) in S. aureus [19].Briefly, duplicate Mueller-Hinton agar plates supplemented with 4% NaCl and 0.5 mg/L oxacillin were incubated in ambient air at 30°C or 35°C for 24 hours.Plates were examined carefully with transmitted light for evidence of small colonies (>1 colony) or a light film of growth indicating oxacillin resistance.Nonstaphylococcal bacteria were not tested for methicillin resistance.Obtained results were compared against phenotypic tests used to support the new assay as well as the QMC phenotypic information held by Dr. Spence.

Statistical Analysis
Negative predictive value (NPV) and positive predictive value (PPV) of the new pentaplex PCR assay were analysed according to the Clinical and Laboratory Standards Institute (CLSI) Guidelines for molecular diagnostics [29].

RESULTS AND DISCUSSION
Use of bioinformatics tools enabled the selection of primer pairs for the amplification of a 174 bp region from ubiquitous bacterial 16S rRNA gene and a 143 bp region from the translation elongation factor Tu (tuf) gene of Staphylococcus spp.Those two primer sets were added into the previously described triplex assay to scale up to the new pentaplex assay thus allowing all the 5 targets to be amplified simultaneously (Fig. 1).
Testing of 377 bacterial isolates on the new assay showed the new assay detected the expected genes from all the bacterial strains studied.The new assay was able to differentiate between Staphylococcus spp.and nonstaphylococcal bacteria.A wide spectrum of S. aureus types, including methicillin-susceptible and methicillin-resistant strains of diverse SCCmec backgrounds previously tested on Dr. Richard Spence's microarray were identified.
As evident in Fig. 2, the pentaplex PCR was able to distinguish staphylococci from nonstaphylococcal bacteria such as E. coli, Proteus spp., and Group A Streptococcus (GAS).The detection of the bacterial 16S rRNA as the only marker in the PCR (Fig. 2, Lane 1) clearly indicated bacterial presence of nonstaphylococcal background.The pattern of Fig. 2 Lane 1 was found with GAS, E. coli, Proteus spp.and other non-staphylococcal bacteria.This is an advantage of the nes assay which is not common to most PCR assays reported in literature.Numerous PCR assays designed to detect S. aureus PVL and MRSA lack the quality of showing the presence of eubacteria of nonstaphylococcal background.This is especially important in pure cultures.Very few such PCRs carry a DNA band for a sequence unique to S. aureus such as the staphylococcal 16S rRNA, a role well played by the staphylococcal tuf in the new pentaplex PCR assay.The detection of the broad-range 16S rRNA would be very useful in the differential diagnosis of skin and soft-tissue infections in which the clinical presentations of staphylococcal and streptococcal diseases are often similar and tend to overlap.As the tuf marker was not detected from nonstaphylococcal bacteria (100% NPV), its detection strongly evidenced the presence of a member of the genus Staphylococcus, thus conferring a double assurance of staphylococcal diagnosis.This is essential both for the institution of empirical therapy and for the purposes of infection control.
Using the new pentaplex PCR assay, gene detection for spa, tuf, 16S rRNA, pvl, and mecA which were combined to inform identification of MRSA, PVL-positive MRSA (PPMRSA), MSSA, PVL-positive MSSA (PPMSSA), methicillin susceptible CoNS (MSCoNS), methicillin resistant CoNS (MRCoNS), and other bacteria corroborated the phenotypic tests used as well as the previous QMC data.All (100%) of the 377 bacteria studied generated the bacterial 16S rRNA marker thus yielding a 100% PPV.All other markers were generated with very high level specificity.No spurious amplification (100% NPV) and no expected marker was left undetected (100% PPV).All other markers were generated with very high level specificity.No spurious amplification (100% NPV) and no expected marker was left undetected (100% PPV).
BLAST data was inadequate in discriminating between PCR fragments of 16S rRNA from bacterial strains of staphylococcal and nonstaphylococcal background.Using the new assay, the staphylococcal tuf showed greater discriminatory power than 16S rRNA.This attribute of the new assay will be helpful in separating clinical isolates of staphylococci (positive for tuf) from other bacteria (negative for tuf).This finding is supported by previous reports [30] and helps to explain the cause of failure of tiny 16S rRNA PCR fragments for taxonomical work as such tiny PCR fragments of 16SrRNA are not discriminatory enough [31].
The new pentaplex PCR assay completes 40 cycles in 37 minutes.The total time from sample preparation to gel documentation, also called the turn-around time (TAT), was 4 hours for a 96-well plate.This is a very good outcome compared with ≥24 hours required to identify staphylococcal methicillin resistance when microbiological agar is in use.This is especially important in blood cultures where the finding of GPC in clusters from positive blood cultures has become a public health concern.
In addition to the PCR primers listed in Table 2, other PCR factors optimized for the new pentaplex PCR assay are summarized in Table 3.Early in the history of PCR, optimization of PCR assays used to be so cumbersome that the list of factors identified as having influence on PCR was never to be complete [32].Though every effort was made to ensure the new assay continues to be useful, we are aware of the new MRSA genomes harbouring the mecC or mecA (LGA251) homologues of mecA.In applying the phenotypic tests to support the new pentaplex PCR assay, we did not encounter strains negative by mecA-PCR and phenotypically oxacillin-resistant, or positive by mecA-PCR and phenotypically oxacillinsusceptible at 0.5 mg/L breakpoint, which is an attribute of the mecA (LGA251) strains [19,33].Since many PCR-based MRSA assays have failed and it is currently perceived that MRSA PCR with only one set of primers for detection of methicillin resistance does not work because of the mecA (LGA251) strains [19,33], we are keeping our eyes on the new pentaplex PCR assay as well as evolving genomes in case any new strain renders the new assay invaluable.

CONCLUSION
Taken together, speed and specificity attributes suggest that the new pentaplex PCR assay is readily adaptable for use in routine diagnostic microbiology laboratories as it will enable the implementation of timely and properly guided therapy and infection control strategies.The only limitation of the new pentaplex PCR assay is the time needed to resolve the PCR amplification products by gel electrophoresis.To exclude the gel resolution time, laboratories with fat budgets and real-time PCR skills can always increase on the speed of the new pentaplex PCR assay by converting the same into real-time PCR since the detectable fragments are all tiny bits which can be possible with real-time PCR.However, the availability of real-time PCR machines capable of simultaneously emitting florescence signal through five channels makes the new assay a good one for laboratories to integrate into their workflow.While we look forward to more realtime PCR machines with ≥5 emission channels and NGS and metagenomic instrumentation capable of sequencing longer templates, the new pentaplex PCR will find use in diagnostic microbiology especially in the developing world where microbiology is known for characteristic slim budgets.